SERA
United States
Enwronmental Protection
Agency
Solid Waste and
Emergency Response
(OS-110W)
EPA/542/N-92/006
December 1992
BiE:.-^-=Jt=I,..1.,--;^,)5g_....J;,.-^ ^f^i-...a..iS*fm
Innovative Measures Distinguish Natural
Bioattenuation from Dilution/Sorption
By John Wilson and D. H. Kampbell, Robert S. Kerr Environmental Research Laboratory
I he EPA's Robert S. Kerr
Environmental Research
Laboratory (RSKERL) recently
demonstrated natural
bioattenuation of organic
contaminants from a gasoline
spill in ground water at the
Sleeping Bear Dunes National
Lakeshore site near Honor,
Michigan. The results of the
demonstration offer conclusive
quantitative evidence of nat-
ural bioattenuation distinct
from dilution and sorption.
Additionally, nitrate depletion,
sulfate depletion and increased
methane production (methano-
genesis) under anaerobic con-
ditions are found to be important
indicators of natural bioattenu-
ation, which should be measured
in addition to the traditional
measure of oxygen depletion.
The geology of the
Sleeping Bear site is charac-
teristic of many urban and
industrial sices where BTEX
compounds (benzene, toluene,
ethylbenzene and xylenes) are
found. The water table aquifer
is in highly transmissive glacial
sands and gravels; and, water
flow is rapid. The ground water
I is cold (10 to 11 degrees
I Centigrade), hard (alkalinity
j 200 to 350 milligrams per liter
(mg/L)) and well-buffered (pH
16.1 to 7.6). Ambient concen-
trations of oxygen, nitrate and
I sulfate in uncontaminated
ground water are 2.4, 67, and
21 mg/L, respectively.
A gasoline service station
had been operated for many
years on a corner lot where
Michigan Highway M-22
crosses the Platte River. The
National Park Service acquired
the land and removed the ser-
vice station. On December 11,
1989, three underground stor-
age tanks were excavated and
removed. Fill and excavated
soil around the tanks smelled
of gasoline. The excavation
was backfilled with the same
soil that had been removed to
prevent injuries to visitors to
the open pit. The distance
from the source of the plume to
the Platte River is only 70 feet.
The following strategy was
used to acquire information
that would allow a quantitative
assessment of natural bio-
attenuation at field scale. A
soil gas survey for hydrocarbon
vapors was used to identify
those areas that still contained
oily phase hydrocarbons and
could act as a source of ground
water contamination. Core
samples were taken in the ar-
eas that showed hydrocarbon
vapors in order to define the
vertical extent of gasoline
contamination and to deter-
mine the depth to which the
spill penetrated moving ground
water. Most of the spill was
above the water table. The
core data and soil gas data
were used to estimate the
total amount of gasoline re-
maining in the subsurface;
(1,200 gallons).
Clusters of monitoring
wells were installed at three
locations in the plume of con-
taminated ground water
at the source of the plume;
at the bank of the river and
at an intermediate point. '
Concentrations of contam-
inants and potential electron
acceptors were monitored i.n
the plume over time.
Attenuation of contami-
nants in ground water plumes
may be due to dilution or
sorption. Unless the contri-
bution by dilution and sorp-
tion can be quantified, it is
impossible to evaluate the
contribution of biodegrada-
tion. One of the goals of this
demonstration was to deter-
mine the extent to which bio-
attenuation is responsible for
the reduction in the concen-
tration of BTEX compounds.
The strategy was to identi-
fy an innocuous component
of the plume that sorbs as
strongly as the contaminants
and that should not be biode-
graded, at least not anaerobi-
cally. The contaminant
plume contained several low
molecular weight branched al-
kanes that should not biode-
grade in the absence of
oxygen, and should sorb at
least as strongly at the BTEX
compounds. One of the al-
kanes, 2,3-dimethylpentane,
was used as a tracer.
For example, in November,
1992, the concentrations of
benzene and toluene in the
(SEE BIOATTENUATION, PACE 3)
Natural Bioattenuation Quantified
DNAPL Guidance
Surfactants Remove DNAPLS
Research and Guidance in Progress
Recycled/Recyclable
Printed with Soy/Canola Ink on paper that
contains at least 5O% recycled fiber
-------
Surfactant Flushing of Ground Water Removes DNAPLs
By John C. Fountain, Ph.D., State University of New York at Buffalo
R,
kecent research results from
a field test have demonstrated
that the surfactant flushing
process is capable of rapid
removal of dense nonaqueous
phase liquids (DNAPLs) from
a contaminated aquifer. The
surfactant solution successfully
removed perchloroethylene
(PCE) at a rate far greater than
it could have been removed by
conventional pump-and-treat
methods. Surfactants have the
ability to greatly increase the
solubility of organic compounds
in water and thus to increase
the efficiency of pump-and-
treat operations. It has been
shown diat pump and treat
alone is not always effective for
restoring aquifers contaminated
with organic compounds.
The research began just
four years ago at bench scale at
the State University of New
York at Buffalo and has been
field tested at the Canadian
Forces Base Borden at Alliston,
Ontario, Canada. It is believed
to be the most carefully
controlled field test of an
aquifer remediation surfactant
flushing process conducted to
date. The field test of surfac-
tant flushing was considered
successful enough to warrant a
pilot scale demonstration at a
contaminated site at Corpus
Chrisd, Texas.
At the Borden field test, a
thrce-meter-square cell was
built in a four-meter-thick
surficial sand aquifer by driv-
ing sheet piling walls into the
underlying clay. A second
sheet piling wall was then in-
stalled one meter beyond the
first wall for secondary con-
tainment. Five injection wells
were installed on one side of
the cell and five extraction
wells on the other side. Ten
multi-level monitoring wells
were also installed.
Next, 231 liters of
reagent-grade PCE were
released into the cell through a
shallow injection well in the
center of the cell. The well
penetrated to a depth of about
30 centimeters (well below the
The pool of DNAPLs,
originally 50 centimeters
thick, was only 3 millimeters
thick at the end of the pilot.
Additionally, out of the 231
liters of PCE injected into the
cell, only 8 liters remained,
spread between the thin layer
above the treatment zone, the
DNAPL pool and in the probe
water table). Thus, the entire
contaminated zone was below
the water table. Prior to
starting surfactant flushing, all
free-phase PCE that could be
directly recovered was pumped
out; approximately 47 liters of
PCE were recovered by this
process. When direct pumping
was no longer recovering
significant volumes of free-
phase PCE, water flushing was
begun to flush any free-phase
PCE that could be so
mobilized and to determine
the vertical and horizontal
variations in hydraulic
conductivity within the cell.
Twelve liters of PCE were
recovered by water flushing.
After recovery of free-
phase PCE by water flushing
ceased, surfactant injection
began. A 2% (by weight)
aqueous solution was used,
composed of equal weights of
nonyl phenol ethoxylate and a
phosphate ester of the nonyl
phenol ethoxylate. The sur-
factant solution containing
the PCE was then pumped
from the extraction wells into
a holding tank, through two
air strippers, into a second
tank where additional surfac-
tant can be added to bring the
surfactant concentration to 2%
if it is necessary to repeat the
surfactant injection again.
Data from the field study
show that the surfactant flush-
ing method circulation of 18
pore volumes would require 90
days of pumping at a standard
rate of 500 gallons per day. By
comparison, it would take
many years to clean the same
volume using conventional
pump-and-treat methods.
Once the free-phase PCE is
removed by surfactant flush-
ing, aquifer restoration can be
accomplished by circulating
water, using a conventional
pump-and-treat operation.
Results from the pilot scale
demonstration at Corpus
_Christi are expected jn_ March
1993. Preliminary data from
surface cores show DNAPLs
below detectable limits. For
more information call John
Fountain at 716-645-3996 at
the State University of New
York at Buffalo. A
ri I I!
JTH E
EPA Directive Updates Ground
Water Policy to Include DNAPLs
By Kenneth Lovelace, Office of Solid Waste
and Emergency Response
CPA's Office of Solid Waste
and Emergency Response
(OSWER) has updated a
1989 ground water policy to
address nonaqueous phase
liquid (NAPL) contaminants,
including dense N APLs
(DNAPLs). The policy
directive promotes a
consistent ground water
cleanup approach for both
Superfund sites and Resource
Conservation and Recovery
Act (RCRA) Corrective
Action sites. The policy
reinforces EPA's commitment
to clean up ground water
contamination at hazardous
waste sites to the fullest
extent possible while
acknowledging at the same
time that, in some situations,
complete ground water
restoration may be technically
impracticable. The directive
builds on previous policies
and good science to address
special problems associated
with NAPL contamination.
It recognizes that DNAPLs
may be more widespread at
hazardous waste sites than
previously realized. The
presence of NAPLs, especially j
(SEE DIRECTIVE, PAGE 3)
-------
B/oattenuation
(from page 1)
spill area were 253 micro-
grams per liter (|Jg/L) and
31,400 (Jg/L, respectively.
The concentration of 2,3-di-
methylpentane was 23.3 |Jg/L in
the spill and 1.6 (Jg/L at the
river bank, 6.9% of the
original concentration in the
spill area. Based on the atten-
uation of 2,3-dimethylpentane,
the expected concentrations
of benzene and toluene due to
dilution or sorption would be
17.4 and 2,170 pg/L, respec-
tively; the actual concentra-
tions were 23.3 and 1.69 pg/L.
A pumping test was con-
ducted to measure the hydrau-
lic conductivity of the aquifer.
The hydraulic gradient was
monitored over time to esti-
mate the direction and veloci-
ty of ground water flow.
Information on flow was used
to predict the average time
elapsed since the water sam-
pled in the well clusters left
the source area of the plume.
The information on time
elapsed and extent of attenua-
tion in ground water was used
to calculate bioattenuation
rate constants.
Bioattenuation of toluene
ranged from 16% to 47% per
week, ethylbenzene from 2.2%
to 7.7% per week, p-xylene
1.7% to 6.7% per week, m-
xylene 1.0% to 2.6% per week
ando-xylene 1.1% to 2.8%
per week. These results are
consistent with those seen in
other methanogenic aquifers
contaminated with petroleum
hydrocarbons. The rate
constants for individual
compounds at sites at Bemidji,
Minnesota, and Traverse City,
Michigan, and at the Sleeping
Bear Dunes site do not vary
more than an order of
magnitude. The agreement is
remarkable, considering the
level of uncertainty intro-
duced into these field scale
estimates from variation in
ground water flow and
changes in plume geometry.
Oxygen consumption, ni-
trate and sulfate reduction,
methane production and iron
solubilization were measured
to estimate the extent of bio-
attenuation of BTEX com-
pounds that could be expected
from these processes. Typically,
as BTEX compounds are de-
graded, there is a decrease in
oxygen followed by anaerobic
degradation supported by ni-
trate and sulfate. After deple-
tion of nitrate and sulfate,
methanogenesis occurs and
methane concentrations in--
crease as BTEX compounds
are further degraded.
Between the spill and the
bank of the river, 42.6 mg/L
BTEX compounds were con-
sumed after correction for
dilution. Methane that accu-
mulated would account for
removal of 39 mg/L of BTEIX,
nitrate reduction for 14 mg/L,
sulfate reduction for 4-2 mg/L,
iron reduction for 1.1 mg/L,
and oxygen respiration for
0.8 mg/L BTEX removed.
The amount of nitrate,
sulfate, oxygen and iron
reduction and methanogenesis
was greater than would be
expected from the quantity of
BTEX compounds removed
from the plume. The actual
electron acceptor demand was
slightly greater than the
theoretical supply of electrons.
Thus, other compounds in the
plume, such as trimethylben-
zenes and naphthalenes, may
have also been biodegraded.
The ground water also
contained large concentra-
tions of non-volatile total
organic carbon, presumably of
natural origin. Total organic
carbon was reduced from 58 mg/L
in the well cluster at the spill
to 47 mg/L 30 feet down
gradient and 21 mg/L 70 feet
down gradient.
Although benzene failed
to degrade, benzene degrada-
_ tionJhas been observed at two
other field sites where the
plume was longer. The failure
of benzene to degrade (after
correction for dilution) at the
Sleeping Bear site is incon-
sistent with these other two
field studies where benzene
was biodegraded. Because the
plume at Sleeping Bear was
short (less than 100 feet) and
the residence time of the
(SEE BIOATTENUATION, PAGE 4)
Directive
(from page 2)
DNAPLs, will significantly
affect the time and likelihood
of achieving ground water
cleanup standards.
Ground water contami-
nation is found at more dian
70% of Superfund National
Priorities List sites and almost
50% of permitted RCRA land
disposal facilities. The direc-
tive emphasizes the need to
determine the likelihood of
NAPL contamination es-
pecially DNAPLs early in
the site investigation. Where
NAPLs are likely, the nature
and extent of contamination
should be characterized to de-
termine appropriate remedial
actions (both early and long-
term actions). Early remedial
Ground Water Currents
actions should be used to
minimize further migration of
dissolved or NAPL contam-
inants. Early response actions
should also include extraction
of free-phase NAPLs, when-
ever possible. Careful ground
water monitoring should be
included in all cleanup efforts
in order to measure effective-
ness and to allow for warrant-
ed design improvements.
Because NAPLs dissolve
slowly, they are a potential
long term source of signifi-
cant contamination.
Accumulations of free-
phase NAPLs, which are not
removed as an early action,
should generally be removed
during the final remedy, to
the extent practicable. The
directive envisions NAPL
remedial actions as part of a
larger phased approach to '
allow coordination with later
cleanup efforts. Furthermore,
because the mass proportion
and spatial extent of residual
NAPLs are usually much
greater than that of free-phase
NAPLs, new conventional
and innovative technologic
should be considered for
enhanced recovery of residual
NAPLs from the subsurface.
For those sites where '
hydrogeologic or contaminant
characteristics may ultimately
make long-term ground water
cleanup targets unattainable,
EPA reserves the right to issue
technical impracticability
waivers for National Priority
List sites and to modify RCRA
permits or enforcement orders.
In such cases, EPA will
identify alternative remedial
requirements, which will
protect human health and the
environment and that are
appropriate for each site's
specific conditions.
For a copy of the full di-
rective, contact die National
Technical Information Ser-
vice at 703-487-4640 and ask
for NTIS Publication No.
PB92-963358, which is
OSWER Directive 9283.1-06
"Considerations in Ground
Water Remediation at Super-
fund Sites and RCRA
Facilities Update." The
directive lists the names and
phone numbers of contacts
at EPA. A
-------
EPA's Multi-Faceted Efforts to Improve Ground
Water Remediation
tPA has several efforts un-
derway to enhance ground
water remediation. The
Robert S. Kerr Environmental
Research Laboratory (RSKERL)
will evaluate innovative tech-
nologies related to ground
water remediation, including
technologies with the potential
to remove nonaqueous phase
liquids (NAPLs) from the sub-
surface. The Office of Solid
Waste and Emergency Re-
sponse (OSWER) will work
closely with RSKERL to de-
velop fact sheets and guidance
on site characterization, reme-
diation and performance mon-
itoring for sites contaminated
with dense nonaqueous phase
liquids (DNAPLs). Additionally,
OSWER has initiated a survey
to determine the potential
number of existing Superfund
sites where DNAPL contam-
ination is likely. This survey,
to be completed by March,
1993, will help to assess the
extent of this problem for the
Superfund program.
OSWER is also support-
ing a National Research
Council (NRG) study, "Alter-
natives for Reducing Risk
from Existing Ground Water
Contamination", that will as-
sess the current state-of-the-
science concerning ground
water remediation and look at
alternative approaches for ad-
dressing ground water con-
tamination. The NRG study
is scheduled for completion by
September, 1993.
Additionally, OSWER re-
cently inventoried alternatives
to pump-and-treat technologies.
For more information on that
study, see the September,
1992, issue of Ground Water
Currents (Document No. EPA/
542/N-92/005).
A technical work group
within OSWER is developing
further guidance concerning
waivers due to technical imprac-
ticability for ground water.
Another work group, led by
the Office of Enforcement, is
developing model consent de-
cree language addressing a
technical impracticability
waiver process for implement-
ed pump-and-treat remedies
at Superfund sites.
Look for news of the
availability of products from all
of these efforts in future issues
of Ground Water Currents. A
Bioattenuation
(from page 3)
ground water was short (5 to 35
weeks), there may not have been
adequate opportunity for anaerobic
degradation of benzene.
For more information, call
John Wilson at EP As Robert
S. Kerr Environmental Research
Laboratory at 405-332-8800. A
To order additional copies of Ground Water Currents, or to be included on the permanent mailing list, send a fax request to the
National Center for Environmental Publications and Information (NCEPI) at 513-891-6685, or send a mail request to NCEPI, 11029
Kenwood Road, Building 5, Cincinnati, OH 45242. Please refer to the document number on the cover of the issue if available.
Ground Water Currents welcomes readers' comments and contributions. Address correspondence to:
Managing Editor, Ground Water Currents (OS-110W), U.S. Environmental Protection Agency,
401 M Street S.W., Washington, DC 20460.
United States
Environmental Protection Agency
National Center for Environmental Publications and Information
Cincinnati, OH 45242
Official Business
Penalty for Private Use $300
BULK RATE
Postage and Fees Paid
EPA
Permit No. G-35
EPA/542/N-92/006
------- |